Journey orchestration platforms enable telcos to separate business logic from core systems, reducing journey deployment time from weeks to days through standardized domains and intelligent orchestration layers that support AI-driven autonomous optimization and real-time adaptation.
Deep Dive
Prerequisite Knowledge
- No data available.
Where to go next
- No data available.
Deep Dive
[In The Loop] Episode 1 – From First Touch to Full Scale - The Power of "Journey Builder"Added:
[music] >> Telcos are actually very good at strategy.
The operators know what they want to launch and they know when they want to launch it.
The frustration we hear from operators is actually about [music] the timing.
By the time the Telco operator ships uh the opportunity has potentially passed.
[music] Uh think about a seasonal offer, uh regulatory change, a reaction to a competitor's move.
>> [music] >> It's always too much effort. Push it through the engineering cycle and operators are waiting weeks for a release window.
Journey Builder gives operators back their timing.
Every journey change, however small, is now a matter of configuration, not engineering effort.
Journey Builder, we believe, reduces the gap between "Here's an amazing idea we want to take to market" >> [music] >> and actually getting that idea live from weeks to days.
The bigger picture though, and that's what really [music] excites me, is that separating the journey logic from your core systems, [music] understanding how they work, controlling the behavior, and optimizing them as you launch into multiple new markets is exactly the prerequisite [music] for our platform to train the AI to start doing this autonomously [music] sometime in the future. And that shift from we need to wait weeks to we can [music] do it just in a matter of days or hours is actually what will set us up for the next big wave of agentic AI.
Because once your journeys are in the orchestration layer, >> [music] >> AI can start operating on them autonomously. Not just helping you configure faster, but actually sensing what's not working and adapting the journey in real time.
We believe that is a foundation for agentic BSS. [music] We envision in the not-too-distant future a system detects a drop-off, proposes a fix, and you [music] approve it in just one click.
The trajectory starts there and these are really, really exciting times.
Honestly, [music] it's fragmented by design, but not by intent.
Most operators are running journeys >> [music] >> stitched together across four or five systems, a CRM, a BSS stack, maybe a campaign tool, and some custom middleware holding it all together.
Each of these systems [music] was built for a different era.
Uh some for voice, some for postpaid, mass market offers.
What you end up with really is a patchwork. And the problem isn't that operators haven't invested. [music] Many have spent significantly. The problem is the architecture was never built for the speed that digital and [music] the segments in consumers of this day and age of AI require.
On the other hand, a journey isn't a campaign.
>> [music] >> Right? It's a continuous dynamic interaction, and most Telco stacks treat it like a batch [music] process.
So, operators are trying to run a real-time business on infrastructure that was designed for quarterly planning cycles.
That is the fundamental tension that we see across every market that we operate in.
Three things consistently.
First, dependency [music] chains.
Changing one step in a journey requires sign-offs and code changes [music] across multiple systems.
There is no single owner, so every change becomes a huge engineering project.
Second, testing environments, or rather the lack of them.
Operators can't safely test the journey modification without touching production, which means changes get batched, delayed, or worse, avoided altogether.
Third, and this one doesn't get talked about enough, organizational misalignment.
The business team decides the journey or the product team decides the journey.
[music] The engineering team builds it and by the time it's live, the market context is already shifted.
You end up launching something that was relevant 6 months [music] ago.
This bottleneck isn't just technical, it's the gap between the people who understand the consumer and the market and the people who control the systems. [music] Johnny Builder was specifically designed to close that gap.
>> [music] >> In most operator environments we've seen, >> [music] >> a small change, reordering KYC steps, adjusting a payment flow, or adding a new onboarding screen takes anywhere from 4 to 12 [music] weeks.
That's not an exaggeration. That is the reality of change request [music] processes, development queues, regression testing, and release cycles.
And this isn't unique to legacy operators. Even digitally ambitious telcos [music] hit this wall because the underlying BSS architecture wasn't built for iteration.
What that means commercially is brutal.
You design an offer for specific acquisition window and by the time the journey is live, that window has passed.
>> [music] >> You are launching into yesterday's market.
>> [music] >> The disconnect is translation loss.
A commercial team has a very clear intent.
Acquire this segment during this window with this offer through this channel.
[music] Simple enough, right? But by the time that intent travels through product [music] specification, engineering tickets, BSS configuration, and multiple testing [music] cycles, it's been interpreted, approximated, and compromised at every single handoff.
What launches is only a version of the original idea and not the idea itself.
And operators have normalized this.
They've built entire planning processes around the assumption that there will be a gap between what the business wants and what the system delivers.
>> [music] >> That normalization is the real problem.
Because it caps ambition.
Right? Teams stop asking for what they actually need because they know that the system can't keep [music] up.
When one thinks of telecom, one thinks of network and router and switches, essentially hardware.
Doing experimentation at hardware level is much more difficult than doing experimentation at the software stack.
[music] Unfortunately, this thinking still prevails in our telecom industry.
It's only in the late '90s and early 2000s with [music] the advent of large internet giants such as Google and Amazon was the concept of quick AB experimentation become really center stage.
Most famously, Google did an experiment to test which of 40 [music] shades of blue gets most clicks on an hyperlink.
Telecom should not be left behind when it comes to experimentation. [music] We at Circles are committed to make sure we change this narrative.
We want to make experimentation [music] a core to everything that we do or that we offer to our customers.
Traditionally, [music] making even a small change in a customer journey used to require weeks of design, coding, testing, [music] launch, monitoring.
Now, imagine you want to do five experiments parallelly.
>> [music] >> It's just a logistical nightmare.
One of our growth [music] product managers observed that many subscribers are not completing their KYC and were just dropping off the funnel.
She wanted to do experiments [music] with sending reminders to the subscribers and see its effect on the conversion rate, what we call CR. [music] She wanted to conduct variety of tests such as whether to send reminders within 24 [music] hours or 48 hours or 72 hours.
With Journey Journey Builder, she was able to set up and test her experiment [music] within hours.
What we're trying to do with Journey Builder is eliminate that translation layer entirely. [music] So, business intent becomes system behavior directly.
When I look at what this team has built over the last 2 years, two bits [music] stand out. First, we standardized our domains. Second, we figured out how to make those domains work together intelligently.
And by solving these problems, what this [music] team has created is now changing the way we build for our mobile network operators.
Telco journeys do not break at the surface, [music] they break underneath.
Any journey, be it onboarding, SIM replacement, plan change, >> [music] >> span across layers like OSS, BSS, network systems, which [music] were not actually designed to work together.
Different vendors, different data models, different assumptions.
The real risk is at the system boundaries. That is where data misaligns [music] and journeys break. And when they do, it just doesn't cause bad customer experience, it also results in double [music] charges, failed activations, or even revenue leakage. The pattern was very clear.
Weak domain boundaries meant fragile orchestration.
And this can't be fixed only at the orchestration layer.
>> [music] >> The foundation had to be fixed first.
The brief was precise. [music] Not clean up the APIs, but the shared language doesn't [music] exist, and that's why everything at the boundaries.
We went domain by domain, identity, billing, [music] provisioning, fulfillment, product catalog, and drew clean boundaries.
>> [music] >> Each domain owns its data, its logic, its APIs. No shared databases, >> [music] >> no shortcuts. Once that was done, we had something powerful and a new problem.
Clean domains that still need to talk to each other for any real customer journey to work.
While the domain work was underway, we focused [music] on the orchestration problem. The requirement was very clear.
State management, intelligent retries, branching, real-time visibility into every journey.
That combination made one thing very clear that this is not a integration or a middleware problem. Instead, this is an orchestration problem.
Once we all aligned on that, the path forward became very clear.
This is an inflection point. The team [music] had defined what the solution needed to look like and build the foundation for that. But the customer doesn't experience domains, [music] they experience journeys.
Those journeys cut across every single domain that had [music] just been cleaned up.
So, how do you compose journeys without undoing the standardization that the team has already done?
That's the question that this team answered by building the journey builder.
Now that the domains were clean, the problem shape [music] was very well understood, and the requirements were precise. The answer to all of them was the conductor.
We followed the mantra, don't stitch inside the domains, instead orchestrate above them.
>> [music] >> What that essentially means is that each domain exposes its capabilities as discrete tasks [music] and conductor owns up the responsibility of stitching them into the journeys.
[music] No domain knows what journey it is participating into. No domain is coupled to one another.
And when [music] we mapped conductor's capabilities against the requirements such as state, branching, intelligent retry, and visibility, conductor delivered all of them. [music] And we were very much impressed with what conductor had to offer to us.
Modifying a journey before conductor.
[music] Let's make it very concrete by taking a real-world example.
Let's say I want to add an identity check [music] of a customer before initiating a payment.
What this earlier meant was that the orchestration logic itself had [music] to change within a service.
You had to get an integration testing done across services and get sign-off from all the teams that were involved. [music] At best, a sprint's effort or even more.
And the overhead still existed for a solution alignment, back-and-forth [music] decision-making among competing priorities.
How conductor helps in modifying a journey.
With conductor, [music] that same change is a new workflow version.
Just define it, test it in isolation, and promote it.
No redeployment, no cross-team release coordination, [music] no change freeze.
This is what low-code actually means from the engineering standpoint. [music] And conductor helps us achieve exactly that.
What we mean is journey logic lives in the workflow definition, not in a deployed code.
>> [music] >> Engineers still build the task and own the domain services, but the composition of those tasks into journeys is now a configuration concern.
>> [music] >> That changes your deployment frequency, risk profile, and operational overhead simultaneously. So, we attacked all the problems with Conductor.
Now, how Conductor helps us win and where the AI enters.
Because Conductor is version controlled, you can run multiple journey versions simultaneously. [music] For example, in an MVNO context, you can have one journey with multiple versions, each catering to different customer needs.
>> [music] >> These are the workflow definitions that coexist, independently testable and promotable as well. Point to notice, these are not the code branches, [music] not even deployment variants, and all these things are achieved using one-stop solution that is Conductor. Now, that brings me to the point of how AI fits into the Conductor.
Treat [music] an AI agent as just another task into the orchestrated journey, where it has its [music] own defined input and defined output. So, what that means is, we are not rebuilding the journey around the AI, the AI already fits into it.
The risk with AI in a customer [music] journey is downstream consequence.
A wrong call at a step three, and steps four and five are already in motion.
>> [music] >> Payment is triggered, SIM is provisioned, or a contract is activated.
You can't easily unwind that.
What Conductor gives us is human review gates. The journey pauses at a designated [music] step, the human reviews the AI's recommendation, the journey proceeds only on [music] explicit approval.
That oversight is the first step in workflow definition.
Not bolted on, but designed in.
>> [music] >> You get AI-assisted decisions where they add value without >> [music] >> removing any human accountability where it matters.
Now, how Conductor helps with observability and reliability. [music] We know that every journey runs through the conductor. Therefore, [music] we have full visibility. For example, which step a customer dropped off or which step had to be replayed. [music] All these things are available at the conductor dashboard. And how this helps developers is that they needn't have to go through n number of services by correlating the logs.
System integrators were always required because every integration [music] was bespoke. A new MVNO always meant one-to-point integrations, [music] new testing cycles, new data mappings.
With standardized domains and conductor [music] in place, we have a pre-built, pre-tested task in stack, which can be used to compose your journey.
>> [music] >> The integrations are already taken care, and what remains is just the configuration.
Process UAE is the proof point. [music] In the previous model, months.
Mapping the KYC requirements, >> [music] >> coordinating the activation sequence across provisioning and billing.
Building custom branching logic, [music] testing, deploying, and stabilizing.
What actually happened? We composed their journey from existing tasks.
UAE [music] specific KYC was a configuration of identity task. Payment sequencing used existing [music] billing capabilities. SIM delivery reused proven provisioning tasks. Months [music] became weeks, and most of that was UAT, not building.
That is the multiplier effect of a [music] task library and the proven orchestration layer.
We diagnosed the problem with enough precision that we knew what we are solving for.
We built the foundation [music] that made everything above it trustworthy.
We eventually built the orchestration layer that lets business move fast on top of it. All these decisions compound, [music] and that's how you build a Telco platform that scales.
A shift that's already underway and will accelerate is from journey execution to journey [music] intelligence.
Today, operators design a journey, deploy and measure it after the fact.
Three to five years from now, >> [music] >> the journey will design itself in real time based on who the customer is, what they're doing and what outcome you're optimizing for. [music] We're moving from static flows to dynamic adaptive experiences.
The other big shift is [music] channel collapse.
Customer support bots are the most celebrated use case of [music] agentic AI around, especially in the telecom.
But let me provide an example from >> [music] >> a customer onboarding journey.
Many telecom operators require their subscribers to upload KYC documents such [music] as driving license for verification.
Only once [music] the verification is completed is the SIM card activated.
With Journey Builder's agentic AI support, >> [music] >> an operator can easy easily plug in an agent to do this verification instead of doing manual verifications. [music] One can even have a human in the loop.
By that I mean if the confidence level of the verification is below a [music] certain threshold, a human can step in.
Journey Builder has been designed to support such and many more agents.
That's what I meant by Journey Builder [music] is truly agentic ready.
Consumers will expect continuity across every touch point and operators that can't deliver [music] that will feel it in Sean.
The underlying requirement if you see for all of this is the same, a configurable headless orchestration layer that can respond to signals in real time.
>> [music] >> That is not a future investment, that's a now investment.
AI agents will do to Telco operations [music] what automation did to manufacturing, but faster and further.
In the near term, we are already seeing agents [music] take over high-volume, rule-based interaction.
Plan recommendations, usage alerts, payment handling, and basic troubleshooting.
That's the [music] surface layer.
The more significant shift is in the back office. AI agents coordinating across [music] systems, managing exceptions, flagging journey drop-offs before they become churn events.
At Circles, we are building toward a model we call Nova.
An operator co-pilot that gives commercial teams real-time visibility and control over journey performance without needing to raise an IT ticket >> [music] >> every time something needs to change.
The operators who will win aren't the ones who deploy the most AI.
It's the ones [music] who integrate AI into the decision layer of their journey orchestration. [music] That's where the real leverage is.
They become infrastructure providers, and infrastructure is a margin compression game.
If you can't move at market speed, you feed the customer relationship to whoever can.
That might be an OTT player, a fintech, a digital native MVNO launching on your own network.
We've already seen this playbook in markets where operators were slow. They end up with the pipes, but not the relationship.
>> [music] >> And the compounding effect is brutal.
Slow systems attract slow talent. Slow talent builds slow road maps, and eventually you're three cycles behind with no clear path back.
The operators we are most worried about aren't the ones with legacy stacks.
[music] Legacy can be worked around.
The ones at real risk are the operators who know they need to change but are waiting for a perfect migration moment that never arrives. In this environment, the cost of waiting is higher than the cost of moving imperfectly.
Journey Builder gives operators something most platforms don't. The ability to separate the business logic of a journey from the underlying systems that execute it.
That sounds architectural, but the commercial implication is very significant. [music] It means an operator can launch a new brand, enter a new segment, or respond to a competitive move without [music] engineering their core stack every time.
The reusability is real. We are targeting 70% reusable components across new market launches.
That's not just faster GTM, that's a structural cost advantage that compounds over time.
And as they layer in AI, experimentation, personalization, performance intelligence, and [music] operator-authored journeys, the platform gets smarter with every deployment.
So, the competitive position isn't just be more faster today.
>> [music] >> It's be learn faster, and that gap widens. For operators thinking about the next decade, Journey Builder is the foundation that makes everything else possible.
>> [music] >> Standardize the domains, then orchestrate them intelligently.
>> [music] >> Both these bets have paid off, and the platform architecture is now ready for AI.
Clean domain [music] are the foundation.
We made sure nothing built on top >> [music] >> could dirty them. You cannot orchestrate a journey in a Telco platform with foundational problems. Identifying that problem and finding the right solution is what made everything [music] else possible. So, all in all, Conductor sits above the domains and composes [music] them.
Journeys evolve without services getting changed. MVNOs get onboarded without rebuilding and [music] AI plugs in without redesigning.
Related Videos
Agentforce NOW AMA: Build with React and Salesforce Multi-Framework
SalesforceDevs
490 views•2026-05-28
How agent o11y differs from traditional o11y — Phil Hetzel, Braintrust
aiDotEngineer
450 views•2026-05-28
WEB TECHNOLOGIES UNIT-2 | Degree 4th sem BCOM Computers web technologies unit-2 full explanation💯✅
LearnwithSahera
1K views•2026-05-29
More tests are always better? How to use AI to identify tests that bring little value
Alliance4Qualification
335 views•2026-05-29
Search Algorithms Explained in 60 Seconds! 🤖💨
samarthtuliofficial
218 views•2026-06-01
People of Game of Thrones using JavaScript DOM
AltCampus
296 views•2026-05-30
Introduction to Problem Solving Part - 1 | Lecture 1 | Intermediate DSA
ascensionix
107 views•2026-05-29
So What's Odin Lang Even Good For
TechOverTea
131 views•2026-06-01











